Device-free sensing techniques that convert CSI amplitude and phase fluctuations into intrusion alerts and gesture classifications, using spatial diversity across multiple receive antennas to localize and characterize human motion.
Explains the extraction of Doppler-shift and amplitude-variance features from CSI time series to distinguish walking, falling, and stationary presence, along with denoising techniques for multipath-rich indoor environments.
Describes how phase differences across receive antennas enable angle-of-arrival estimation for device-free localization, and how multiple access points fuse CSI streams for room-level intrusion coverage.